Comparing Likelihood Ratios to Understand Genome-Wide Variation in Phylogenetic Support

Area of Science:

• Evolutionary Biology
• Genomics
• Computational Biology

Background:

• Phylogenomic studies can produce conflicting evolutionary trees despite robust support for competing hypotheses.
• Understanding variations in phylogenetic signal across datasets is crucial for resolving these conflicts.

Purpose of the Study:

• To compare maximum and marginal likelihood ratios for evaluating phylogenetic conflicts.
• To investigate how different likelihood ratio calculations affect conclusions about squamate evolutionary relationships.

Main Methods:

• Calculated and compared maximum and marginal likelihood ratios across multiple phylogenomic datasets.
• Examined support on a locus-by-locus basis to understand data set variation.
• Evaluated recalcitrant relationships among major squamate lineages.

Main Results:

• Likelihood ratios show broad correlation across loci, but this is driven by extreme values.
• The choice of likelihood ratio method and data filtering can change the proportion of loci supporting a hypothesis.
• Maximum likelihood ratios often provide identical support for alternative hypotheses, complicating conflict resolution.

Conclusions:

• The interpretation of phylogenomic conflicts depends on the likelihood ratio method used.
• This study provides surprising evidence for a sister relationship between snakes and iguanians across four datasets, challenging previous findings.